Table 3.
Multivariate analyses of the factors associated with aGvHD (grades I–IV).
| Entire group, n = 99 | ||||
| Parameters | Number of CD4+ lymphocytes (×103 cells/L) | Percentage of CD4+CD25high lymphocytes | Number of CD4+CD25high lymphocytes (×103 cells/L) | MUD donor |
| Coefficient | 0.001 | 1.394 | −0.529 | 0.963 |
| p value | 0.496 | 0.223 | 0.031 | 0.047 |
| Odds ratio | 1.001 | 4.031 | 0.589 | 2.619 |
| −95% Cl | −0.003 | −0.863 | −1.009 | 0.014 |
| +95% Cl | 0.005 | 3.651 | −0.049 | 1.911 |
| In vivo T cell-depleted group (patients receiving ATG or Campath), n = 73 | ||||
| Parameters | HLA mismatch | Anti-CMV IgG serostatus (D−/R+) | Number of CD4+CD25high lymphocytes (×103 cells/L) | MUD donor |
| Coefficient | 0.600 | −0.423 | −0.363 | 1.970 |
| p value | 0.347 | 0.487 | 0.048 | 0.029 |
| Odds ratio | 1.822 | 0.655 | 0.696 | 7.173 |
| −95% Cl | 0.515 | 0.196 | 0.485 | 1.224 |
| +95% Cl | 6.450 | 2.194 | 0.997 | 42.016 |
* Table 3 shows the results of the forward stepwise logistic regression analysis. During the first step analysis, we included factors that had already been suggested [21,23,24,36,40,41,42] to contribute to the risk of aGvHD, including the type of donor (sibling or unrelated), the level of donor-recipient HLA matching (9/10 or less vs. 10/10 and sibling-matched transplantations), CMV reactivation events during the first year post-HSCT, age, anti-CMV-IgG serology (donor serology and donor/recipient serology: negative/positive), mode of transplantation (MAC vs. RIC), transplant material (peripheral blood progenitor cells (PBPCs) vs. bone marrow (BM)), the female-to-male transplantation, percentage and the numbers of CD4+ and CD4+CD25high lymphocytes.